An embedded variable rate speech encoding scheme having scalability in the signal band is attracting attention as an speech encoding scheme capable of supporting from conventional call services to active wideband speech communication services. Further, since scaleable encoding information is such that encoding information can be freely reduced at arbitrary nodes on the transmission channel, it is effective in congestion control in communication utilizing packet networks typified by an IP network. As a result of this background, band-scaleable embedded variable rate encoding schemes of speech signals are subject to standardization in ITU-T (International Telecommunication Union—Telecommunication standardization sector) SG16 (Study Group 16).
On the other hand, in speech signal encoding, LSP parameters are widely used as parameters for effectively representing spectrum envelope information and LSP parameter encoding is also one of essential, elemental technologies in band-scaleable speech encoding.
When the LSP parameters are to include band scalability, wideband LSP parameters are subjected to predictive quantization by using narrowband LSP parameters obtained by analyzing narrowband signals. Therefore, prediction accuracy and quantization efficiency in predictive quantization of wideband LSP parameters are important indicators directly influencing band scaleable encoding performance of speech signals.
As technology for performing predictive quantization of wideband LSP parameters such as these, technology is also well known (for example, refer to Patent Document 1) for predicting wideband LSP parameters from encoded narrowband LSP parameters by using non-linear prediction technology such as codebook mapping, generating the prediction difference by comparing these prediction results with actual wideband LSP parameters, and transmitting both the generated prediction difference and encoded narrowband LSP parameters. Further, technology is also well-known (for example, refer to Patent Document 2) for predicting wideband LSF parameters from narrowband LSF (Line Spectral Frequency) parameters using, for example, codebook mapping and encoding prediction residuals.    Patent Document 1: Japanese Patent Application Laid-open No. 2003-534578.    Patent Document 2: Japanese Patent Application Laid-open No. Hei6-118995.